I have two internal 1TB SATA drives (not RAIDed), one from Seagate, one from Western Digital.
I use them as storage drives and they're usually unplugged. When I need to copy something over, I do it by hand to both.
Occasionally I use TestPath (freeware, lightweight, but possibly buggy utility) to compare that everything is the same.

On Monday, after one such file compare, I did a file search on the WD and upon reboot, chkdsk started running on it. This is on Windows 2000 SP4 (all patched up).
Looking at the event viewer, minutes before the reboot, there was this entry.
System log. Error (red). Type: ntfs.


I only encountered this once before, and it was while running testpath; this was the time I tried to copy all data from a large 500 GB drive to both drives (then new). In that situation both drives were trashed and I decided to dban them and redo the copy and compare (which went fine - the 500 is retired now).

I ran chkdsk manually on the Seagate and it was fine.

QUESTION 1: Can a utility (read-only I think) corrupt an ntfs file system? Or do you think it was the controller or a hardware fault?


To keep it short, I got a new PC during the week (kind of because of the error, and also because I'd been putting it off - see my old thread about bad power and computer failure). Installed XP SP3 this time, drivers, updates, and everything. I plugged in the two drives, and while searching the Seagate, I get a popup about the ntfs file system being corrupt! Surely enough, the event viewer gives the same message, but for drive S: now. Being careful, I disabled chkdsk from running on that drive at boot time


chkdsk in read-only mode gives errors and quits.



QUESTION 2:
How could chkdsk have decided the file system was fine (I'd ran it on the Seagate after the WD was corrupt), then decide several days later that the drive is corrupt?


To be continued... I'll post questions about interpreting chkdsk results in a separate thread.